ADT has become increasingly popular over the past decade as primary therapy for older men with newly diagnosed localized disease not receiving other curative intent treatments (surgery or radiotherapy), and as a treatment for a rising PSA after failure of these curative therapies. In both groups, there is no proven mortality benefit from clinical trials, and it is unlikely that trials will be conducted. Given the increasing number of elderly men, the high incidence and survival rates from prostate cancer, and the use of ADT in an estimated 100,000 men with prostate cancer each year, there is urgent need for information on outcomes to inform treatment decisions. We propose to assess fatal and morbid outcomes up to 15 years following androgen deprivation therapy (ADT) for men initially diagnosed with localized prostate cancer. In addition to lack of evidence of a mortality benefit, there is emerging evidence suggesting that ADT may be associated with increased cardiovascular mortality, as well as musculoskeletal (bone fractures), cardiovascular, and endocrine-related (diabetic) events. Our three specific aims include estimating the benefits and risks of ADT in terms of all cause and prostate-cancer specific mortality, estimating the rate of serious non-fatal adverse events; and assessing whether commonly used prescription medications can lessen the harms associated with ADT. We will assess cancer-specific outcomes according to prognostic risk groups defined by age, stage, serum biomarker values (PSA), and other pathological markers of tumor aggressiveness. For adverse event estimates, we will account for variations in baseline comorbidity and clinical predictors, and will use state-of-the-art effectiveness analysis techniques to correct for selection bias. The retrospective observational study will be conducted using a large, diverse population of over 45,000 men diagnosed from 1995-2007 with a mean follow up of 6 years. There are comprehensive computerized clinical utilization data for this population from 3 large integrated health care plans, including longitudinal information on tumor characteristics, risk factors and outcomes. Key variables will be derived from inpatient, outpatient, pharmacy and radiology data and lab test values. In contrast to prior observational studies, ours will have the combination of size, follow up, and detailed clinical information over the entire disease trajectory needed to significantly improve the precision of risk estimates associated with ADT. These strengths, and our multi-disciplinary team experienced in prostate cancer outcomes research using large databases, will yield new and important information needed to improve treatment decision making and outcomes. PUBLIC HEALTH RELEVANCE: Hormonal therapy for prostate cancer has become increasingly popular over the past decade. However, there is no proof that this therapy can prolong survival from the disease, and there are several potentially serious long-term unintended consequences from its use. These adverse effects include bone fractures, cardiovascular disease, and diabetes. Randomized controlled trials are not planned to evaluate these issues. Given the ageing of the population, and the use of hormonal therapy in an estimated 100,000 new men each year, there is an urgent need for better information on outcomes to inform treatment decisions. In our proposed study we will quantify the mortality benefits and adverse effects of hormonal therapy for prostate cancer. To perform this study we will use a detailed clinical database containing extensive information on 45,000 men diagnosed in 1995-2007 in three large health plans. Our results will provide new information to help men make more informed decisions about starting hormonal therapy.